Confirm whether the data points in the matrix were entered as row vectors or column vectors. A vector is a matrix that has either a row of length one or column of length one. Return to the documentation or original data that accompanies the matrix. This information should make it clear whether the data points were treated as row vectors (1 by n in dimension, where "n" represents the number of variables) or column vectors (n by 1 in dimension). If you do not have this information, the best way to make this determination is to check the similarity among rows and columns. If the numbers in the rows look more similar to one another than the numbers in the columns, the data points are column vectors and the rows represent variables; otherwise, the data points are row vectors and the columns represent variables.
Check whether there is a coefficient acting as a variable. For some matrices, such as those representing mathematical functions, a row (for column vector data points) or column (for row vector data points) may represent a coefficient, which is not a real variable. This coefficient will either be at the beginning of the matrix or the end of the matrix. If the documentation or understanding of the matrix leads you to believe there is a coefficient present, delete the row or column corresponding to the coefficient from the matrix.
Cut the matrix by variable. If the data points are entered as row vectors, cut the matrix into separate columns. If the data points are entered as column vectors, cut the matrix into separate rows.